Unraveling the Hidden Dynamical Structure in Recurrent Neural Policies

Recurrent neural policies are widely used in partially observable control and meta-RL tasks. Their abilities to maintain internal memory and adapt quickly to unseen scenarios have offered them unparalleled performance when compared to non-recurrent counterparts. However, until today, the underlying mechanisms for their superior generalization and robustness performance remain poorly understood. In this study, by analyzing the hidden state domain of recurrent policies learned over a diverse set of training methods, model architectures, and tasks, we find that stable cyclic structures consistently emerge during interaction with the environment. Such cyclic structures share a remarkable similarity with \textit{limit cycles} in dynamical system analysis, if we consider the policy and the environment as a joint hybrid dynamical system. Moreover, we uncover that the geometry of such limit cycles also has a structured correspondence with the policies' behaviors. These findings offer new perspectives to explain many nice properties of recurrent policies: the emergence of limit cycles stabilizes both the policies' internal memory and the task-relevant environmental states, while suppressing nuisance variability arising from environmental uncertainty; the geometry of limit cycles also encodes relational structures of behaviors, facilitating easier skill adaptation when facing non-stationary environments.

E2PL: Effective and Efficient Prompt Learning for Incomplete Multi-view Multi-Label Class Incremental Learning

Multi-view multi-label classification (MvMLC) is indispensable for modern web applications aggregating information from diverse sources. However, real-world web-scale settings are rife with missing views and continuously emerging classes, which pose significant obstacles to robust learning. Prevailing methods are ill-equipped for this reality, as they either lack adaptability to new classes or incur exponential parameter growth when handling all possible missing-view patterns, severely limiting their scalability in web environments. To systematically address this gap, we formally introduce a novel task, termed \emph{incomplete multi-view multi-label class incremental learning} (IMvMLCIL), which requires models to simultaneously address heterogeneous missing views and dynamic class expansion. To tackle this task, we propose \textsf{E2PL}, an Effective and Efficient Prompt Learning framework for IMvMLCIL. \textsf{E2PL} unifies two novel prompt designs: \emph{task-tailored prompts} for class-incremental adaptation and \emph{missing-aware prompts} for the flexible integration of arbitrary view-missing scenarios. To fundamentally address the exponential parameter explosion inherent in missing-aware prompts, we devise an \emph{efficient prototype tensorization} module, which leverages atomic tensor decomposition to elegantly reduce the prompt parameter complexity from exponential to linear w.r.t. the number of views. We further incorporate a \emph{dynamic contrastive learning} strategy explicitly model the complex dependencies among diverse missing-view patterns, thus enhancing the model's robustness. Extensive experiments on three benchmarks demonstrate that \textsf{E2PL} consistently outperforms state-of-the-art methods in both effectiveness and efficiency. The codes and datasets are available at https://anonymous.4open.science/r/code-for-E2PL.

DSAEval: Evaluating Data Science Agents on a Wide Range of Real-World Data Science Problems

Recent LLM-based data agents aim to automate data science tasks ranging from data analysis to deep learning. However, the open-ended nature of real-world data science problems, which often span multiple taxonomies and lack standard answers, poses a significant challenge for evaluation. To address this, we introduce DSAEval, a benchmark comprising 641 real-world data science problems grounded in 285 diverse datasets, covering both structured and unstructured data (e.g., vision and text). DSAEval incorporates three distinctive features: (1) Multimodal Environment Perception, which enables agents to interpret observations from multiple modalities including text and vision; (2) Multi-Query Interactions, which mirror the iterative and cumulative nature of real-world data science projects; and (3) Multi-Dimensional Evaluation, which provides a holistic assessment across reasoning, code, and results. We systematically evaluate 11 advanced agentic LLMs using DSAEval. Our results show that Claude-Sonnet-4.5 achieves the strongest overall performance, GPT-5.2 is the most efficient, and MiMo-V2-Flash is the most cost-effective. We further demonstrate that multimodal perception consistently improves performance on vision-related tasks, with gains ranging from 2.04% to 11.30%. Overall, while current data science agents perform well on structured data and routine data anlysis workflows, substantial challenges remain in unstructured domains. Finally, we offer critical insights and outline future research directions to advance the development of data science agents.

Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning

In this report, we introduce DASD-4B-Thinking, a lightweight yet highly capable, fully open-source reasoning model. It achieves SOTA performance among open-source models of comparable scale across challenging benchmarks in mathematics, scientific reasoning, and code generation -- even outperforming several larger models. We begin by critically reexamining a widely adopted distillation paradigm in the community: SFT on teacher-generated responses, also known as sequence-level distillation. Although a series of recent works following this scheme have demonstrated remarkable efficiency and strong empirical performance, they are primarily grounded in the SFT perspective. Consequently, these approaches focus predominantly on designing heuristic rules for SFT data filtering, while largely overlooking the core principle of distillation itself -- enabling the student model to learn the teacher's full output distribution so as to inherit its generalization capability. Specifically, we identify three critical limitations in current practice: i) Inadequate representation of the teacher's sequence-level distribution; ii) Misalignment between the teacher's output distribution and the student's learning capacity; and iii) Exposure bias arising from teacher-forced training versus autoregressive inference. In summary, these shortcomings reflect a systemic absence of explicit teacher-student interaction throughout the distillation process, leaving the essence of distillation underexploited. To address these issues, we propose several methodological innovations that collectively form an enhanced sequence-level distillation training pipeline. Remarkably, DASD-4B-Thinking obtains competitive results using only 448K training samples -- an order of magnitude fewer than those employed by most existing open-source efforts. To support community research, we publicly release our models and the training dataset.

FasterPy: An LLM-based Code Execution Efficiency Optimization Framework

Code often suffers from performance bugs. These bugs necessitate the research and practice of code optimization. Traditional rule-based methods rely on manually designing and maintaining rules for specific performance bugs (e.g., redundant loops, repeated computations), making them labor-intensive and limited in applicability. In recent years, machine learning and deep learning-based methods have emerged as promising alternatives by learning optimization heuristics from annotated code corpora and performance measurements. However, these approaches usually depend on specific program representations and meticulously crafted training datasets, making them costly to develop and difficult to scale. With the booming of Large Language Models (LLMs), their remarkable capabilities in code generation have opened new avenues for automated code optimization. In this work, we proposed FasterPy, a low-cost and efficient framework that adapts LLMs to optimize the execution efficiency of Python code. FasterPy combines Retrieval-Augmented Generation (RAG), supported by a knowledge base constructed from existing performance-improving code pairs and corresponding performance measurements, with Low-Rank Adaptation (LoRA) to enhance code optimization performance. Our experimental results on the Performance Improving Code Edits (PIE) benchmark demonstrate that our method outperforms existing models on multiple metrics. The FasterPy tool and the experimental results are available at https://github.com/WuYue22/fasterpy.

Sketch-in-Latents: Eliciting Unified Reasoning in MLLMs

While Multimodal Large Language Models (MLLMs) excel at visual understanding tasks through text reasoning, they often fall short in scenarios requiring visual imagination. Unlike current works that take predefined external toolkits or generate images during thinking, however, humans can form flexible visual-text imagination and interactions during thinking without predefined toolkits, where one important reason is that humans construct the visual-text thinking process in a unified space inside the brain. Inspired by this capability, given that current MLLMs already encode visual and text information in the same feature space, we hold that visual tokens can be seamlessly inserted into the reasoning process carried by text tokens, where ideally, all visual imagination processes can be encoded by the latent features. To achieve this goal, we propose Sketch-in-Latents (SkiLa), a novel paradigm for unified multi-modal reasoning that expands the auto-regressive capabilities of MLLMs to natively generate continuous visual embeddings, termed latent sketch tokens, as visual thoughts. During multi-step reasoning, the model dynamically alternates between textual thinking mode for generating textual think tokens and visual sketching mode for generating latent sketch tokens. A latent visual semantics reconstruction mechanism is proposed to ensure these latent sketch tokens are semantically grounded. Extensive experiments demonstrate that SkiLa achieves superior performance on vision-centric tasks while exhibiting strong generalization to diverse general multi-modal benchmarks. Codes will be released at https://github.com/TungChintao/SkiLa.

Use of Continuous Glucose Monitoring with Machine Learning to Identify Metabolic Subphenotypes and Inform Precision Lifestyle Changes

The classification of diabetes and prediabetes by static glucose thresholds obscures the pathophysiological dysglycemia heterogeneity, primarily driven by insulin resistance (IR), beta-cell dysfunction, and incretin deficiency. This review demonstrates that continuous glucose monitoring and wearable technologies enable a paradigm shift towards non-invasive, dynamic metabolic phenotyping. We show evidence that machine learning models can leverage high-resolution glucose data from at-home, CGM-enabled oral glucose tolerance tests to accurately predict gold-standard measures of muscle IR and beta-cell function. This personalized characterization extends to real-world nutrition, where an individual's unique postprandial glycemic response (PPGR) to standardized meals, such as the relative glucose spike to potatoes versus grapes, could serve as a biomarker for their metabolic subtype. Moreover, integrating wearable data reveals that habitual diet, sleep, and physical activity patterns, particularly their timing, are uniquely associated with specific metabolic dysfunctions, informing precision lifestyle interventions. The efficacy of dietary mitigators in attenuating PPGR is also shown to be phenotype-dependent. Collectively, this evidence demonstrates that CGM can deconstruct the complexity of early dysglycemia into distinct, actionable subphenotypes. This approach moves beyond simple glycemic control, paving the way for targeted nutritional, behavioral, and pharmacological strategies tailored to an individual's core metabolic defects, thereby paving the way for a new era of precision diabetes prevention.

TALENT: Table VQA via Augmented Language-Enhanced Natural-text Transcription

Table Visual Question Answering (Table VQA) is typically addressed by large vision-language models (VLMs). While such models can answer directly from images, they often miss fine-grained details unless scaled to very large sizes, which are computationally prohibitive, especially for mobile deployment. A lighter alternative is to have a small VLM perform OCR and then use a large language model (LLM) to reason over structured outputs such as Markdown tables. However, these representations are not naturally optimized for LLMs and still introduce substantial errors. We propose TALENT (Table VQA via Augmented Language-Enhanced Natural-text Transcription), a lightweight framework that leverages dual representations of tables. TALENT prompts a small VLM to produce both OCR text and natural language narration, then combines them with the question for reasoning by an LLM. This reframes Table VQA as an LLM-centric multimodal reasoning task, where the VLM serves as a perception-narration module rather than a monolithic solver. Additionally, we construct ReTabVQA, a more challenging Table VQA dataset requiring multi-step quantitative reasoning over table images. Experiments show that TALENT enables a small VLM-LLM combination to match or surpass a single large VLM at significantly lower computational cost on both public datasets and ReTabVQA.

Interference Alignment for Multi-cluster Over-the-Air Computation

One of the main challenges facing Internet of Things (IoT) networks is managing interference caused by the large number of devices communicating simultaneously, particularly in multi-cluster networks where multiple devices simultaneously transmit to their respective receiver. Over-the-Air Computation (AirComp) has emerged as a promising solution for efficient real-time data aggregation, yet its performance suffers in dense, interference-limited environments. To address this, we propose a novel Interference Alignment (IA) scheme tailored for up-link AirComp systems. Unlike previous approaches, the proposed method scales to an arbitrary number $\sf K$ of clusters and enables each cluster to exploit half of the available channels, instead of only $\tfrac{1}{\sf K}$ as in time-sharing. In addition, we develop schemes tailored to scenarios where users are shared between adjacent clusters.

FastAvatar: Towards Unified Fast High-Fidelity 3D Avatar Reconstruction with Large Gaussian Reconstruction Transformers

Despite significant progress in 3D avatar reconstruction, it still faces challenges such as high time complexity, sensitivity to data quality, and low data utilization. We propose FastAvatar, a feedforward 3D avatar framework capable of flexibly leveraging diverse daily recordings (e.g., a single image, multi-view observations, or monocular video) to reconstruct a high-quality 3D Gaussian Splatting (3DGS) model within seconds, using only a single unified model. FastAvatar's core is a Large Gaussian Reconstruction Transformer featuring three key designs: First, a variant VGGT-style transformer architecture aggregating multi-frame cues while injecting initial 3D prompt to predict an aggregatable canonical 3DGS representation; Second, multi-granular guidance encoding (camera pose, FLAME expression, head pose) mitigating animation-induced misalignment for variable-length inputs; Third, incremental Gaussian aggregation via landmark tracking and sliced fusion losses. Integrating these features, FastAvatar enables incremental reconstruction, i.e., improving quality with more observations, unlike prior work wasting input data. This yields a quality-speed-tunable paradigm for highly usable avatar modeling. Extensive experiments show that FastAvatar has higher quality and highly competitive speed compared to existing methods.